Transcript editorby CadreScripts

Your projects

Signed outSign in to save your work and transcribe

Local only

What are you working on?

Pick a starting point. You can switch between transcript, captions and translation at any time.

Transcribe a recording

Upload audio or video. You get a transcript with speakers, timestamps and words to check.

Choose a fileNeeds sign-in

Edit a transcript you have

Open a Word, text or HTML file. It exports exactly as it came in unless you change it.

Import document.docx .txt .html

Make captions for a video

Import .srt or .vtt, or build captions from a transcript. Reading speed and line length are checked for you.

.srt .vtt .ttml .scc

Add an audio or video file to start

Drop a recording here, open one from your device, or paste a video link. You can also import a transcript to edit.

Open mediaImport document

Signed out, nothing leaves this device. Press ? any time for keyboard shortcuts.

00:00:00

TimelineDrag to move · edges to retime Zoom20 s Wave×1 Height

Make captions

Captions are built from your transcript and its timings, made by transcription (signed in), or imported from a caption file (SRT, WebVTT, TTML, DFXP, SCC or SBV). Every caption can then be edited, styled and timed here.

Import a caption file

Publishing or broadcasting? Get human-verified captions from CadreScripts ↗

Transcript Captions

From

Proofread

Click a name in the transcript to change who is speaking.

More speaker tools

Find speakers automatically when a document is uploaded Recognise names as I type (“Name: ” at the start of a paragraph)

Tidy merges one- or two-word switches that interrupt a sentence (a common diarization slip) into the surrounding speaker.

Status

no project

Comments

Hover a paragraph (or a row in Translation) and click the comment button beside it to comment on that place. Ctrl+Alt+C comments where the cursor is.

Attach player time

Share this project

People

Editors change the text and send it for review. Reviewers can’t edit; they comment, then approve or ask for changes. People need an account on the editor; your team is listed in Account → Your team.

Names and terms for this project, so spellings stay consistent.

Word or name	Note

Anything you paste here becomes plain text. Copy a term and paste it into the transcript and it takes on the transcript’s formatting. Pasting a list adds one row per line.

GlossaryNames and terms kept consistent

Checks

English spelling

US English UK English

Spelling mistakes are also underlined in the text.

Settings

Check the number style (three, 25, 5%, $20) Suggest commas and full stops

AutoCorrect

Bookmark a paragraph with the bookmark button in the toolbar (or right-click the transcript); bookmarks also show on the player’s timeline. Save clips from the right-click menu: they download in the recording’s own format (signed in).

History needs an account. Right now your work is kept only in this browser, as one latest copy, so there are no earlier versions to go back to. Clearing browser data or switching devices loses it, so export a copy to keep it safe.

to save to the cloud every minute and keep your last 7 versions.

No account yet? Accounts are created by our Admin. Email us or use the contact page.

Transcription Language Also spoken

For speakers who switch languages: those parts are transcribed in the language spoken, not translated.

Identify speakers

Transcript style VerbatimEvery word as spokenUm, so we, we installed them last month. Clean verbatimWithout um, uh and repeatsSo we installed them last month. Intelligent verbatim (smooth sentences)Edits shown as suggestions you acceptWe installed them last month.

Apply the number style (one to nine in words; 10 and above, % and $ in figures)

Names and termsNone yetSpelled your way in Proofread and translations

This recording

Open a recording, or paste a video, Drive or Dropbox link, to see what you can do with it here.

Your plan

Sign in to see your plan. Transcription, accurate timestamps and translation count toward it; editing, captions and exports are free.

Don’t have an account? Contact us or use the contact page, and we’ll set one up for you.

Without signing in, your work stays on this device.

Connection settings

Help

Keyboard shortcuts: , or press ? outside the text.

✨ What’s new

Upgrade to human-verified: in a project, Review → Human-verified (or the offer in the sidebar): choose a transcript, captions or translation check and a turnaround, see the price and the share of your plan it uses, and send it. The checked version comes back as its own project; Show changes marks every word our team changed, and it exports with the changes (Word with tracked changes, PDF, text).
Episode kit (Episode tab, for podcasts): title ideas, a hook, the description, chapters with exact times (ready for YouTube), takeaways, quotes, names mentioned, keywords and social posts, made from the transcript and saved with the project.
Caption compliance audit (Captions → Compliance audit): which videos have no captions or only automatic ones, in your projects, a YouTube channel or a list, with a dated report for ADA Title II and WCAG 2.1 AA.
Engines: transcription runs on AssemblyAI, with ElevenLabs taking over automatically if a file fails; translation runs on Claude. Free tools: SRT ⇄ VTT, timing shifter, reading speed checker, cost calculator, timestamps.
Change case: select words (or click in a word) and press Aa after Highlight, or right-click → Change case: Sentence case, lowercase, ALL CAPS or Capitalize Each Word. Shift+F3 goes round them, as in Word.
Capitals and punctuation: the first word of a sentence gets its capital as you type (after a full stop, at the start of a paragraph and after a speaker’s name). Proofread finds turns and paragraphs that start with a small letter, “, And” in the middle of a sentence, and missing commas and full stops, each with Fix.
Text size: click the percentage beside A− and A+ to choose a size from a list, as in Word.
Lists and indents: bulleted and numbered lists and indents, after Highlight in the toolbar; in a list, Tab and Shift+Tab move an item in and out. Word files with lists keep their look.
Bookmarks: the bookmark button after Find bookmarks the paragraph you are in and lists your bookmarks and clips; bookmarks also show as marks on the player’s timeline (click one to go there).
Glossary: in the Transcribe panel (add the names before you transcribe), at the top of Proofread, and right-click a word or name → Add to glossary.
Keyboard shortcuts: All shortcuts under the text, or press ?; they are in groups, with a search box.
The player sits under the text; for the floating window, choose Floating player window in your menu.

Questions about your plan or account? Contact us at info@support.cadrescripts.com.

This editor is made by CadreScripts. When accuracy matters, our team can do the work for you: human-verified transcription and video captions.

🔒 Privacy & Data Storage

Not Signed In (Local Session): Your imported media, transcripts, and documents stay on your device. Your transcript, glossary and shortcuts are stored in your browser’s local storage and are not uploaded to our servers (media files are not stored, so open them again after a reload), and transcription is only available when you are signed in. If you paste a YouTube or Vimeo link, that service’s player loads in your browser. Browsers can clear local storage, so export anything you want to keep.
Signed In (Cloud Session): Your work automatically syncs to the cloud every minute when it has changed. History keeps one version for every 5 minutes of work, plus one each time you press Save; we retain your last 7 working versions, transcript and captions together, so you can restore prior edits while the project exists. Each project’s glossary is saved with it. Recordings you open stay on your computer by default: only transcripts, captions and edits go to the cloud. A recording is uploaded to your account’s cloud storage when you press Transcribe (it is sent to AssemblyAI for transcription, or to ElevenLabs if you choose it or if AssemblyAI cannot do the file; generating accurate timestamps for an uploaded transcript sends it to Amazon Transcribe), when you choose “Save to cloud” on it, when you order human-verified work (a CadreScripts agent then works on a copy of the project and listens to the recording), or always if you pick that under Account. A project whose recording stayed on a computer asks for the file when you open it again.
Two-step sign-in: Account → Two-step sign-in → Turn on, then scan the QR code with an authenticator app on your phone (Google Authenticator, Microsoft Authenticator, 1Password, Authy or similar) and type the 6-digit code it shows. From then on, each sign-in asks for a fresh code from the app after your password, so someone who learns your password still cannot get in. To turn it off you need your password and a code. Lost your phone? Contact us: an administrator can turn it off for you.
Wider panels: Drag the inner edge of the projects sidebar (left) or of the panel beside the icons (right: Proofread, Speakers, Review, Episode…) to make it wider or narrower; the ⤢ button in the panel’s header makes it wide in one press. Double-click an edge for the usual width again. The editor always keeps enough room, and your widths are remembered on this computer. With the keyboard: Tab to the edge, then the arrow keys (Shift for bigger steps), Home, End, or Enter for the usual width.

📁 Media & Project Limits

Storage Allocation: Audio and video uploads share a total account storage limit of 5 GB, unless your plan includes more (the bar under Your projects shows yours).
Video Files: Up to 3 GB per file (the same limit applies to audio), or what your plan includes. Videos are automatically deleted after 30 days, like the rest of the project. Audio uploads faster and uses less of your storage, so upload audio whenever you can. Transcription accepts files up to 2 GB and 8 hours.
Links (signed in): Paste a Google Drive, Dropbox, OneDrive or direct file link (shared with “Anyone with the link”) and press Transcribe: the audio is downloaded into your storage first, then transcribed as usual. YouTube and Vimeo do not allow downloading audio, so upload the file instead; for videos on your own YouTube channel, use “Import YouTube captions”.
Your projects: The list is in folders by age: New (the first 2 days), then Week 1 to Week 4 (deleted soon). Recordings transcribed together are in a batch folder inside them. Tick projects, a folder or a batch to Export them together into one .zip (Word, PDF, text, SRT, WebVTT and the recordings) or to delete them. A project’s ⋯ button has Open, Recording (downloads its recording when it is stored in the cloud; also Export → Download the recording, for the open project), Export and Delete. The button beside the search opens the project library, every folder on one page.
Batch processing: Under New project, Batch processing transcribes several recordings at once (up to 50): pick or drop them, name the batch, choose the language, speakers and transcript style, and see the cost before anything starts. Each recording becomes its own project in one batch folder. You can keep working, or close the page: the transcripts go into their projects as they finish (also after a reload).
Deleting your account: Account → Delete account. Type delete to confirm. Your projects and recordings are deleted at once and everything else linked to your account within 30 days, so download what you want to keep first; your login stops working and we are told.
Project Retention: Cloud projects are automatically removed 30 days after creation. Always export copies of transcripts and glossaries you wish to keep permanently.

🎧 Playback & Interactive Sync

Jump to Audio: Click any timestamp in the editor (or any transcribed word) to jump straight to that point in the recording. Click or drag on the waveform to move the playhead; the coloured bar under it shows who speaks when.
Speakers: With speaker labels on, each new speaker starts a paragraph with their name. Rename, merge or delete speakers, or reassign a paragraph, in the Speakers panel, or by clicking a name in the transcript (“Change this label only” changes just that turn). The speaker shortcuts (Ctrl+Shift+1, Ctrl+Shift+2…) put “Speaker 1:”, “Speaker 2:”… where the cursor is. Deleting a speaker who is in the transcript asks first, and keeps their words. Typing “Name: ” at the start of a paragraph makes it that speaker’s label (Ctrl+Z keeps it as plain text); typing a speaker’s first letters suggests the name, and Tab completes it.
Smooth playback: If a video doesn’t play smoothly, a bar offers Light playback (free, for everyone): the recording plays as sound only, so your device does less; Show the video brings the picture back. Signed in, your plan can include a smoother copy: a smaller copy (480p) of the video made in the cloud, played instead of the original (you see its share of your plan first). Exports and burned-in videos still use the original. Playback in the file strip opens these choices any time.

✏️ Editing Tools

Right-click menu: Right-click the transcript: the top of the menu shows the moment you clicked (its time, the speaker, the recording's waveform with the part a clip would take in amber, and the words you selected). From there, play from that word, save a clip (the selected words, or the paragraph) or download it (signed in; in the recording's own format: MP4, MOV, M4A, MP3, AAC and WAV clips are cut from the file in moments, without losing quality; WebM, MKV and other formats are recorded as the clip plays, as WebM for WebM files), bookmark a paragraph, comment, highlight in five colours (the last chip, the eraser, removes a highlight), strike through, add a word or name to the glossary, offset the timestamps, undo, redo, copy, cut, delete, or find and replace. A greyed-out item says why when you point at it. Highlights and strikethrough go into Word and PDF exports. Bookmarks and clips have their own tab inside Review. Shift+right-click shows the browser’s own menu (spelling suggestions). Keys: Ctrl+Shift+Space plays from the cursor, Ctrl+Shift+H highlights, Ctrl+J strikes through (change them in Shortcuts).
Caption animation and colours: In the Captions panel → Style, choose how captions come on (fade, slide, pop, word by word as they are said, or the spoken word bounces), on the player and in burned-in videos. In Speakers, tick Colour each speaker’s captions and pick a colour per speaker; SRT, WebVTT and TTML exports can include the colours.
Brand kit (signed in): Save your caption style, brand colours and logo once, and apply them to any project in one click; the logo can sit in a corner of the video. Turning it on takes a small one-time fee from your plan for 90 days; afterwards your kit is kept until you turn it on again.
Videos with captions burned in: Export → Video with captions burned in is for signed-in accounts.
Translation: Signed in, choose Translation (or Translate… in the toolbar’s Open menu) to translate the transcript or the captions into one or more of 75 languages. Each translation is kept with the project: review it side by side with the original, edit it, mark paragraphs as reviewed, undo and redo changes (Ctrl+Z / Ctrl+Y), translate one row again in place, export it (also side by side), or open it as its own project. The Admin chooses the engine for each language (Amazon Translate, or Claude Haiku or Sonnet on Amazon Bedrock) and can compare them on a real transcript.
Transcript styles: Verbatim keeps every word. Clean verbatim removes filler words (um, uh) and words repeated by accident. Intelligent verbatim (smooth sentences) uses AI in the cloud, so it is for signed-in users; it also removes false starts (“I think it’s, um, I think it’s the same” → “I think it’s the same”), stutters and self-corrections, and can fix small grammar slips one word at a time. These changes are never made silently: they appear as suggestions, struck through in red with any new word in green, like track changes. Click one to accept or reject it, step through them with the bar above the transcript, or use Accept all / Reject all in Proofread. Words that stay keep their timing (click-to-play), speaker labels and timestamps. Uploaded transcripts can be tidied the same way.
Number style: Proofread checks numbers: one to nine in words, 10 and above in figures, percentages and dollar amounts in figures with % and $ (“five percent” → “5%”, “twenty dollars” → “$20”). Numbers that start a sentence, times, dates, ages, scores, measurements and labels such as “Chapter 3” are left alone. Fix or Fix all; the check can be switched off. Transcribe → Apply the number style writes new transcripts that way.
Layout checks: Proofread lists double spaces, empty paragraphs (double paragraph breaks), a paragraph that appears twice and the same speaker on two turns in a row, each with Fix or Ignore. Nothing is changed until you choose.
AutoCorrect: In Proofread → AutoCorrect, keep a list of words to replace as you type (for example “teh” → “the”, or “xx” → “[inaudible]”). Ctrl+Z straight after a correction keeps what you typed. Capitalize the first letter of sentences (on unless you switch it off) gives the first word of a sentence its capital when you type the space or punctuation after it: after a full stop, question mark or exclamation mark, at the start of a paragraph, and after a speaker’s name or a timestamp there. Not after “Mr.”, “e.g.” or “…”, and not for words such as iPhone or x.com; Ctrl+Z straight after puts back what you typed.
Change case: Select words (or click in a word) and press Aa after Highlight in the toolbar, or right-click → Change case, then choose Sentence case, lowercase, ALL CAPS or Capitalize Each Word. Shift+F3 goes round them, as in Word. Only the letters change: bold, highlights, links and the words’ times stay, and speaker labels and timestamps are left as they are. Sentence case keeps “I”, glossary names and the speakers’ names as they are written; Capitalize Each Word writes glossary terms as the glossary does (iPhone). One Ctrl+Z puts it back.
Capitals and punctuation in Proofread: Checks lists a turn or paragraph that starts with a small letter (not a paragraph that carries on the sentence before it, or list items), a capital in the middle of a sentence such as “there, And then” (Fix makes it small; Full stop ends the sentence before it instead), missing commas (after “However”, “Of course”, “Yes” or “Well” at the start of a sentence, before “but” or “so” that joins two sentences, and before the name of someone spoken to: “Thanks, Sophie.”) and a missing full stop at the end of a turn or paragraph (a question mark when it reads as a question, with the other mark as a second button). Speech is not always written with every comma, so these are suggestions; switch the commas and full stops off in Proofread → Settings.
Text size: A− and A+ in the toolbar make the text smaller or larger; the percentage between them opens a list of sizes (70% to 200%), as Word’s size box does. Ctrl+Alt+= larger, Ctrl+Alt+− smaller, Ctrl+Alt+0 back to 100%. Transcript, Captions and Translation each keep their own size. It is how big the text looks on your screen: exports and uploaded documents keep their own font sizes.
Player: Once a recording is open, the player sits under the text: the video (⤢ makes it larger), skip back, play, skip forward, the time, the speed and the waveform with who speaks when. The arrow beside the skip buttons shows the other skip lengths; the one you choose is what F2 and F3 use. Prefer a window? The pop-out button on the player (or Floating player window in your menu) floats it over the editor: drag its bar to move it, an edge or corner to resize it, minimise it to its bar, or dock it again. Captions mode with a video puts the video beside the captions, so it never covers them.
Your menu: The button beside your name in the sidebar opens Account and plan, Your team, the Admin console (administrators), Keyboard shortcuts (or press ? when you are not typing), Theme, Floating player window, Help and privacy, Contact us and Sign out. The panel tabs on the right show the tools for the mode you are in.
Starting out: With nothing open, the editor shows three ways in: transcribe a recording, edit a transcript you have, or make captions for a video. “Try it with a sample transcript” opens a short two-speaker interview to try things on. Typing in the editor below them, or opening anything, starts working straight away.
Other languages in a recording: Under the gear beside Transcribe, “Also spoken” names a second language. Parts spoken in it are transcribed in that language, as spoken, not translated.
Links in the text: Select words and press the link button beside B, I and U (or Ctrl+K), then type or paste a web or email address; with nothing selected it adds a new link at the cursor. Click a link to open it in a new tab, change its address or remove it (the words stay); Ctrl+click opens it straight away. Links from an uploaded Word document work the same way. Word and PDF exports keep the links; plain text keeps the words.
Lists and indents: The list buttons after Highlight make the paragraphs you are in a bulleted or numbered list, and move them in or out (Decrease indent, Increase indent); on a narrow screen they are under one list button. In a list, Tab and Shift+Tab move an item in and out a level, Enter on an empty item ends the list, and Backspace at the start of an item takes it out of the list. Shortcuts (change them in Keyboard shortcuts): Ctrl+Shift+8 bullets, Ctrl+Shift+7 numbers, Ctrl+M and Ctrl+Shift+M indent. Word, PDF and text exports keep the lists and indents (Word as real Word lists).
Find & Replace: Use the Find button in the toolbar (or Ctrl/Cmd+F while your cursor is in the transcript) to search your transcript.
Timestamps: The Timestamp button (or its shortcut, F4) inserts one at the cursor; selected words stay. With the cursor on a timestamp, or before the speaker of a line that starts with one, it gives that timestamp the current time instead (Ctrl+Z puts the old time back). To put one before a speaker with the mouse, click the speaker's name and choose Timestamp at the start (or press F4 while that menu is open). Click a timestamp to play from it; Alt+click (Option+click on a Mac) only puts the cursor there. Type over its digits to change its time. Backspace just after a timestamp (or Delete just before it) removes it whole. A timestamp you type yourself, such as [00:01:23], works as one once you type the space after it. The arrow beside the button opens Timestamps: Generate timestamps adds them for you, every 30 seconds, every minute, every paragraph, or at a Custom… spacing you type (in minutes or seconds, for example every 2 or 5 minutes). Each one goes at the start of the paragraph nearest its mark, so a long paragraph keeps a single timestamp, at its start. “Accurate, from the audio” needs an account; the estimate works for everyone. The format, layout (on its own line, or at the start of the text) and brackets are in the same place; they apply to new timestamps, and Apply to the document changes the ones already there (timestamps in “Name (00:00):” speaker labels only when you tick them). Generating again replaces the earlier generated timestamps; Remove generated keeps any words you typed on their lines.
Review: Words the transcription was unsure about are underlined. The Review bar steps through them and plays each one.
Exports in any language: Word keeps every language exactly. PDF shows Korean, Chinese, Japanese, accented letters (Vietnamese, Turkish, Polish…), Greek, Russian and other Cyrillic, Armenian, Georgian and Amharic; the first such PDF downloads a font (a few seconds). Languages whose letters join or change shape (Arabic, Farsi, Urdu, Hebrew, Hindi and other Indian languages, Thai) go to Word instead: the editor tells you before anything is made. In Translation → Export you choose whether timestamps are included, on their own line or at the start of the text, their format and brackets.
Share a view-only link: In Review → Share this project, choose Create link. Anyone with the link can read the transcript and play the recording, but can’t edit or comment. You choose whether they may download it, and you can turn the link off or replace it at any time. It works until the project is deleted.
Tags (F9): In the transcript, F9 opens a menu of tags such as [inaudible 00:12:34] and [crosstalk 00:12:34], with the playhead time filled in; press 1–9 to pick one. Choose “Edit the list…” to make your own (saved with your settings).
Pause while typing: Under Shortcuts, “Pause the recording while I type” stops the recording as you type and plays it again, from a little earlier, a moment after you stop.
Title block: In Export → Document, tick “Title block” to start Word, PDF and text files with the title, recording, length, speakers and date. Translations add the language.
Checking a translation: Rows that may need a second look are marked: a number missing or changed, different end punctuation, a fixed translation or name not used, text left untranslated, italics lost, or a caption much longer than the original. The Checks filter shows only those rows; Ignore hides a mark you have checked. The copy button on a row puts the original into it (for names, numbers, or to translate it yourself). Export → PDF table gives time, original and translation side by side. Caption translations keep italics and places on screen, are wrapped to the line length for the new language and checked against its reading speed.

🎬 Captions

Rules: In the Captions panel → Rules, choose a rule set (Netflix adult programmes, Netflix children’s programmes, US broadcast or Custom), then Subtitles or SDH and the language. Each language gets its own limits (for example 16 characters a line in Chinese and Korean, 42 in English). You can change the line length, reading speed, shortest and longest time, the gap between captions in frames, chaining and snapping to video frames; changing one makes the rules Custom, and Reset goes back.
Checks: The Captions panel → Checks lists errors and warnings (the selected caption shows its own in its row, with Fix and Ignore): lines too long, reading too fast, too short or too long on screen, gaps and overlaps, chaining gaps, shot changes, a new speaker without a name or dash, line breaks that could fit on one line, the same text twice, and italics or brackets not closed. Fix mends one, Ignore hides it for that caption, and Fix all re-wraps, cuts captions that are too long, closes tags and fixes the timing. Fix is offered only where it clears the check: for reading speed or a caption that is too short it uses free time around the caption or spare time of its neighbours (moving a time by at most a second), and joins a caption too short for its words to the next one of the same speaker. When no time is free (fast speech, captions back to back), shorten the text, set the times by hand, raise the reading speed in Caption rules if your guidelines allow it, or Ignore the warning. Undo with ↶.
Styles and places: Select words and press I, B or U (Ctrl+I, Ctrl+B, Ctrl+U), or with nothing selected the whole caption. Place puts a caption at the top, middle or bottom, left or right. Italics, bold, underline and places are kept when you import and export SRT, WebVTT, TTML/IMSC and DFXP; SCC keeps italics, underline and the row on screen.
Formats: Export and import SubRip (.srt), WebVTT (.vtt), TTML / IMSC (.ttml), DFXP, Scenarist SCC (.scc, for broadcast: 32 characters a line, and letters the broadcast standard cannot show are named before you export) and YouTube SBV (.sbv). Every export is read back to make sure it opens as it should.
Several captions at once: Ctrl/Cmd-click to pick captions one by one, Shift-click for a run, Ctrl+A for all (in a caption’s text, Ctrl+A first selects its words and pressed again selects every caption). Merge, Delete (or the Delete key) and Timing then work on the ones picked; Deselect in the toolbar (or Esc) unpicks them without deleting anything.
Timing: Timing → move times earlier or later (seconds or frames, all or the picked captions), change the frame rate (for example 23.976 to 25), or sync with two points (put the first and last caption where they belong; the rest follow). Times can be shown with frames (00:01:02:12).
Shot changes: Timing → Find shot changes looks through the video in your browser (nothing is uploaded) and marks the cuts on the timeline. Captions then snap to them when you drag, the checks follow Netflix’s shot-change rules, and “Snap captions to shot changes” moves the times for you.
Sound tags: Tags (or F9) adds [laughter], [applause], ♪ around the selected words and more, for SDH captions. Edit the list to make your own.
Comments on a caption: Signed in, the speech-bubble button (Ctrl+Alt+C) comments on the caption you are in. Captions with comments show a mark, and Review lists them with the caption number.
Burned-in captions: Export → Video with captions burned in makes a video file with the captions drawn on the picture, in your browser, so it is free and nothing is uploaded. It takes as long as the video plays, so keep the tab open. The file is MP4 where the browser can make one (Chrome, Edge, Safari), otherwise WebM.
Publish to YouTube: Signed in with YouTube connected, Export → Publish captions to YouTube puts the captions on one of your videos, replacing the track in the same language if there is one.
How captions look on the video: Captions panel → Style (Larger opens it beside a big preview), as in an editing program’s properties panel: presets (Classic, Netflix, Yellow, Social, Band, Minimal), the video shape (the video’s own, landscape 16:9, portrait 9:16, square 1:1 or portrait 4:5, with the picture fitted inside or cropped to fill), font, weight, size, capitals, colour, line spacing, background (a box behind each line, one box, a full-width band or none) with its colour and opacity, outline and shadow, the distance from the edge, safe zones for Reels, TikTok and Shorts, and highlighting the spoken word. The size is a share of the picture’s shorter side, so captions on a portrait video fit it. The preview shows the frame exactly as a burned-in video will be; over the player the captions sit on the picture itself.
Timeline: Drag its bottom edge (or use Height, or Alt+wheel) to make it taller; Wave makes the waveform taller (Ctrl+Shift+wheel); Zoom shows more or fewer seconds (Ctrl+wheel). When the video is paused, the time under the cursor shows: Ctrl+C copies it, and right-click copies it (as shown, as an SRT time or in seconds), starts or ends the selected caption there, adds a caption there or moves the playhead. While the video plays, the timeline turns pages instead of scrolling every frame, so long or large videos play smoothly.
On a phone: The text comes first, with a small player above the tabs (tap the picture for a larger video). In Captions mode the video stays at the top as you scroll, and the tools for the selected caption scroll sideways.

⌨️ Shortcuts & Customization

Customize your keyboard hotkeys and timestamp triggers anytime under Shortcuts. Transcript and Captions modes each have their own keys; playback keys work in both.